Objectives: To determine the incidence of various causes of sudden unexpected death (SUD) within an entire population and to assess the relative importance of an expert autopsy, as well as age of demise, in predicting the likelihood of finding a cause of death.
Methods: We reviewed all cases of SUD in infants aged 1 week to 18 months that occurred in the province of Quebec (Canada) between 1987 and 1996.
Results: We identified 623 cases of SUD; in 80% the diagnosis was sudden infant death syndrome (SIDS). Infection was the most common non-SIDS diagnosis (7.1% of all SUDs), followed by cardiovascular anomalies (2.7%), child abuse or negligence (2.6%), and metabolic or genetic disorders (2.1%). The percentage of non-SIDS deaths was much higher for autopsies performed in centers with expertise in pediatric pathology (18% vs 6%, P <.005). The likelihood of a non-SIDS diagnosis was much higher at age ranges atypical, as compared with typical, for SIDS (33% at 7 to 27 days, 19% at 6 to 12 months, and 64% at 12 to 18 months [atypical] vs 15% at 1 to 6 months [typical]; P =.003).
Conclusion: The study of an entire population provides more reliable data regarding causes of SUDs than does the study of small groups. We recommend that in addition to a thorough investigation of each SUD, autopsies be performed in centers with expertise in pediatric pathology. This recommendation takes on added significance in this era of decreasing SIDS rates.